A new study using both genetic and cultural data shows that ethnic groups in Central Asia are primarily a sociocultural phenomenon.
From the Abstract:
…we used genetic data that … in addition to data from the literature, to understand better the origins of Central Asian groups at a fine-grained scale, and to assess how ethnicity influences the shaping of genetic differences in the human species. We assess the levels of genetic differentiation between ethnic groups on one hand and between populations of the same ethnic group on the other hand …
Our results show that there are more differences between populations of the same ethnic group than between ethnic groups for the Y chromosome, whereas the opposite is observed for mtDNA in the Turkic group. This is not the case for Tajik populations belonging to the Indo-Iranian group where the mtDNA like the Y-chomosomal differentiation is also significant between populations within this ethnic group. Further, the Y-chromosomal analysis of genetic differentiation between populations belonging to the same ethnic group gives some estimation of the minimal age of these ethnic groups. This value is significantly higher than what is known from historical records for two of the groups and lends support to Barth’s hypothesis by indicating that ethnicity, at least for these two groups, should be seen as a constructed social system maintaining genetic boundaries with other ethnic groups, rather than the outcome of common genetic ancestry.
Our analysis of uniparental markers highlights in Central Asia the differences between Turkic and Indo-Iranian populations in their sex-specific differentiation and shows good congruence with anthropological data.
To the extent that ethnicity can be seen as parallel to, equal to, or nascent racial groupings, this is evidence supporting the model that for the most part distinct racial groups are genetic unicorns even if they are hard and stark realities in the political and social world. This is not to say that genetic variation does not exist or that it is not (somewhat) meaningful, especially with respect to understanding historical factors. But, the race-based model of correlated alleles that allow an observer to place a person in an ethnic group and then predict invisible features (like IQ, criminality, math ability, etc) is brought into serious question.
The study concludes:
Since the work of Frederik Barth in the 1970s  anthropologists have placed emphasis not only on presumed common ancestry and shared cultural traits, but also on the “boundaries” used by individuals in order to distinguish themselves from members of other ethnic groups. These boundaries can take different forms – racial, cultural, linguistic, economic, religious, and political – and may be more or less porous. The persistence of such boundaries implies rules. One of the most common rules around the world is an endogamous preference for mate choice. In conclusion, our analysis of uniparental markers lends support to Barth’s hypothesis by indicating that ethnicity, at least for two (and marginally three) of the Turkic groups in Central Asia, should be seen as a constructed social system maintaining genetic boundaries with other ethnic groups rather than the outcome of common genetic ancestry.
Barth F: Ethnic groups and boundaries. The social organization of culture difference. (Results of a symposium held at the University of Bergen, 23rd to 26th February 1967.). Bergen, London: Universitetsforlaget;Allen & Unwin; 1969.
Heyer, E., Balaresque, P., Jobling, M., Quintana-Murci, L., Chaix, R., Segurel, L., Aldashev, A., & Hegay, T. (2009). Genetic diversity and the emergence of ethnic groups in Central Asia BMC Genetics, 10 (1) DOI: 10.1186/1471-2156-10-49